Overview

Scikit-learn 1.7, released on June 15, 2025, improves the HTML representation of estimators and adds Poisson loss to MLPRegressor.

Main Features

Improved HTML representation

Estimators now display an interactive HTML representation in Jupyter notebooks, making pipeline inspection easier.

python 
from sklearn.pipeline import Pipeline
from sklearn.preprocessing import StandardScaler
from sklearn.svm import SVC

pipe = Pipeline([
    ('scaler', StandardScaler()),
    ('svc', SVC()),
])
pipe  # HTML representation in Jupyter

MLP with Poisson loss

The MLPRegressor now supports Poisson loss, suitable for count data.

python 
from sklearn.neural_network import MLPRegressor
from sklearn.datasets import make_regression

X, y = make_regression(n_samples=500, random_state=42)
y = abs(y)  # positive values for Poisson

mlp = MLPRegressor(loss='poisson', max_iter=500)
mlp.fit(X, y)
print(f'Score: {mlp.score(X, y):.3f}')

Sources